One of the most challenging aspects of plant breeding is to identify plant varieties that are superior to the currently available varieties used in commerce. Herein, the term “variety” and “genotype” will be used interchangeably since genetic differences are what make each variety unique and what make one variety superior to another in terms of commercial value.
For commodity crops like soybeans and corn, the most universal measure of commercial value is grain productivity per unit area or “yield”. Since a farmer is paid according to the quantity (weight) of grain he delivers to an elevator, a farmer typically wants to plant a variety that produces the most grain per acre.
Although yield is arguably the most important trait that a plant breeder is concerned with, it is also the least understood genetically. There are many different plant traits that control the efficiency of converting nutrients and light into grain. Yield is therefore the final culmination of many different traits that contribute to productivity over the growing season. These would include seedling emergence vigor, photosynthetic ability, disease resistance, ability to mine nutrients from the soil, ability to produce flowers, and ability to shuttle photosynthate into grain, etc. The genetic bases of these individual traits that contribute to yield are largely unknown. Each trait that contributes to “yield” could be controlled by several or many genetic loci. Therefore, the overall genetic basis for yield is undoubtedly very complex. This is just one reason why traditional methods of determining the genetic basis of yield have not been very successful. To make incremental improvements in yield potential, for the most part, plant breeders are still using the same resource intensive methods that have been in use for the last 80 or more years. Existing varieties are crossed to produce an array of new genotypes which are then exhaustively tested over many locations and replications in order to get enough yield data to differentiate the few consistently superior genotypes. This is one of the most expensive and time-consuming aspects of plant breeding.
During the 1990's, genetic markers linked to genes that contribute to yield emerged as a means to improve efficiency in certain aspects of the breeding process. These success stories have been limited to traits that are controlled by relatively few genes that are highly heritable. In this case, it is fairly routine to make a reliable association between a DNA sequence and a phenotype that can be confirmed with a greenhouse or field assay. However, until very recently, it has been extremely difficult to make reliable associations between specific DNA sequences and a very complex quantitative trait such as yield.
“Breeding bias,” described in U.S. Pat. No. 5,437,697, which is incorporated herein in its entirety for all purposes, is a unique way to determine which genetic loci have been affected by extended periods of recurrent selection for yield. By comparing the genetic marker profiles of modern high yielding varieties to their most distant ancestors, breeding bias can quickly leverage an entire century of yield data to determine which specific alleles of which genetic markers have increased in frequency over time due to selection. Since increased yield has been the main criteria for selection, these markers are those most likely to be associated with yield progress over time. The present invention provides genetic markers that are associated with yield performance in a variety of geographic regions, as well as methods for utilizing these markers to efficiently identify soybean lines and sublines with increased yield.